abcamara.com
Applied #IR & #NLP Research @zeta-alpha.bsky.social, making search good. CS PhD @tudelft. Former @naverlabseurope.bsky.social, @bloomberg.com CNF✈️AMS
👤 Joined on 01 July 2023
Followers
Following
Total Posts
@acamara.bsky.social@acamara.bsky.social → @abcamara.comIn many ways, it feels like the epistemics of the models are getting way worse with increased RL -- they're super eager to solve, not eager to understand. Some older model, like GPT-4.5 and pre RL large models had a much better world model. I'm cautiously hoping Fable 5.5 helps solve some of this.
Show PostExploring my sexuality in my stand up special Big Divorce Energy! Watch it for free on Youtube October 23!
Show Postreading articles critical of LLM coding that start with, "Using Claude Sonnet 5" and I just have to stop. There's 2 classes of models: Frontier and Lenny. Sonnet is a Lenny model. Lenny only gets to write if Frontier wrote the instructions. Otherwise, sit in the corer and eat paste, Lenny.
Show Post